A multi-species benchmark for training and validating mass spectrometry proteomics machine learning models
preprint · ChemRxiv · 2024
preprint · ChemRxiv · 2024. Bo Wen et al. Training machine learning models for tasks such as de novo sequencing or spectral clustering requires large…
| Date | 2024-08-30 |
| Type | preprint |
| Venue | ChemRxiv |
| Publisher | ChemRxiv |
| Contribution | benchmark |
| DOI | 10.26434/chemrxiv-2024-z5b8m |
| Citations (OpenAlex) | 1 |
Peer-reviewed version: A multi-species benchmark for training and validating mass spectrometry proteomics machine learning models (2024-11-08, Scientific Data)
Abstract
Training machine learning models for tasks such as de novo sequencing or spectral clustering requires large collections of confidently identified spectra. Here we describe a dataset of 2.8 million high-confidence peptide-spectrum matches derived from nine different species. The dataset is based on a previously described benchmark but has been re-processed to ensure consistent data quality and enforce separation of training and test peptides.
Methods and tools
- 9-species multi-species benchmark: Multi-species PSM benchmark for ML training
Cited by (1)
- Deep Learning Methods for De Novo Peptide Sequencing (2024) crossref